In　order　to　analysis　the　impact　of　vehicle　condition　on　its　safety　states,　a　fuzzy　comprehensive　assessment　method　was　put　forward　on　the　basis　of　fault　tree　analysis.　The　fault　tree　of　vehicle　safety　states　was　established　on　the　basis　of　scientific　analysis　of　influencing　factors　of　vehicle　safety　states.　The　sorting　of　structure　importance　of　all　the　basic　events　was　obtained　by　calculating　the　structure　importance.　Based　on　the　fault　tree　analysis　result　of　vehicle　safety　states,　nine　key　elements,　namely　loading　weight,　speed　and　horizontal　safety　distance　etc,　were　picked　out　and　considered　as　the　elevation　indexes　to　construct　the　fuzzy　comprehensive　assessment　model　of　vehicle　safety　states　based　on　fuzzy　evaluation　theory.　The　developed　vehicle　monitoring　and　early-warning　system　that　could　real-time　monitor　the　operating　states　parameters　of　the　tested　vehicle　was　installed　in　a　CA1046L2　light　truck.　The　road　tests　were　conducted　according　to　the　pre-set　fault　test　method.　The　safety　states　of　the　tested　vehicle　were　evaluated　using　the　above-mentioned　assessment　model.　The　test　results　showed　that　the　model　for　vehicle　safety　states　assessment　had　good　applicability.